Image: Adobe Stock

Alternative data: exploring the ethical implications

Tue Feb 25, 2020
$download_content = get_field('download_content');

We’ve worked with our partners at Refinitiv to explore ‘alternative data’ – data from new and unique sources used by investors to help evaluate a company or investment –  to understand what data is being used and how it is being collected

Alt data has recently boomed in popularity, opening many new avenues for data inputs into investment models

In our paper with our partner Refinitiv, Building an Open and Trustworthy Alt Data Ecosystem, we explore this topic in depth and look at how to improve standardisation and ethics in this emerging space.

Read the Chinese translation here (PDF). With thanks to Refinitiv for the translated version. Chinese translation: Building an open and trustworthy alternative data ecosystem. 构建开放和可信的 “ 另类数据 ”
生态系统

Those working in finance and investment circles may know what alternative data is or at least have heard of it as the latest buzzword to hit your office. For everyone else, ‘alternative data’ (or ‘alt data’) is data used by investors to help evaluate a company or investment that comes from new and unique sources.

We want to build an open and trustworthy data ecosystem. So, in our research we have looked across a variety of models that increase access to data while maintaining trust, to maximise value and mitigate harm. This project is of special interest to us as we can look at the inner workings of an emerging data ecosystem.

To learn more about alternative data we’ve worked with Refinitiv to explore this growing trend and understand what data is being used, how it is being collected, and what steps can be taken to build an open, trustworthy and sustainable alternative data ecosystem.

Alt data encompasses a wide variety of datasets, including social media sentiment, credit card transactions, satellite imagery, and almost everything in between. Over the course of our research we have discovered some very interesting uses of data that have both piqued our interest and led us to consider their ethical implications.

For example, some companies are analysing real-time sentiment data, in the form of positive or negative social media posts, and making predictions on company performance before any traditional data is released from the company. Other businesses are using corporate aviation data, including flight patterns of private jets, to try and uncover meetings between companies and predict the next big merger or acquisition. One company, JetTrack, claims their solution discovered Amazon’s acquisition of Whole Foods. While SpaceKnow are using satellites to monitor manufacturing capacity in China to create an index they claim is even more accurate than the Chinese government.

As you might imagine, people and organisations are becoming concerned about these activities and their privacy and ethical implications. People may be happy with personal data being used for benefits to society, as noted in a recent poll, but may not want that data being used to assist the investment decisions of hedge funds.

The alt data ecosystem is also suffering from a lack of standards. It includes vast amounts of unstructured data that takes time, talent, and technology to properly analyse. This means that the quality, provenance and limitations of the datasets being sold in the market today is often not known by alt data users until this investment has been made. Transparency helps us monitor collection and use of data and helps downstream consumers of these products decide whether or not they want to use them based on their source.

In our paper we’ve suggested some ways to address these ethical, technical and practical issues. Tools like our ‘Open Standards for Data’ handbook can help drive adoption of standards, while our Data Ethics Canvas can help organisations identify and make decisions about potential ethical issues. Assuming that data can be ethically collected and reused, then tools like PermID can help with data integration.

We’re planning to do more work to explore how to assess open, trustworthy data ecosystems, so if you want to discuss the report, or work with us in other emerging areas then get in touch [email protected]